Professional Documents
Culture Documents
0022-2380
When will individuals speak up about organizational issues, and when
will they remain silent? We suggest that organizational voice will be significantly
influenced by individuals’ perceptions of the attitudes towards an issue within their
workgroup. In particular, individuals will be more likely to speak up when they
believe that their position is supported by others, and remain silent when they believe
that it is not. We explain this using the ‘spiral of silence’ proposed by Noelle-
Neumann (1974, 1985, 1991) and widely used in public opinion research, which
explains how majority opinions become dominant over time and minority opinions
weakened.
Spirals of silence within groups can restrict the open and honest discussion that
is essential to organizational improvement. Noelle-Neumann’s spiral of silence
emphasizes the horizontal pressures that the threat of isolation and corresponding
fear of isolation exert to keep people from being open and honest about their
opinions. We argue in this paper that the fear and threat of isolation are particularly
powerful for members of invisible minorities such as gay and lesbian employees. We
propose a second, vertical ‘spiral of silence’ may develop through processes at a more
micro level within the workgroup and organization. This second spiral begins with
the inability to fully express one’s personal identity within the workgroup because of
a negative climate of opinion towards a particular aspect of one’s identity. This may
be especially true for ‘invisible’ sources of diversity such as sexual orientation.
Revealing a potentially disruptive identity might impair social cohesion: concealing
it, however, can inhibit social exchange and task exchange and reduce self-efficacy,
leading to organizational silence. However, an alternate virtuous spiral can result in
which individuals will feel empowered to express organizational voice.
INTRODUCTION
Contributions from employees are essential to empowerment and more open lines
of communication (e.g., Lawler, 1992; Pfeffer, 1994; Spreitzer, 1995), and for
Address for reprints: Frances Bowen, Haskayne School of Business, University of Calgary, 2500
University Drive NW, Calgary, Alberta, Canada T2N 1N4 (Frances.bowen@haskayne.ucalgary.ca).
© Blackwell Publishing Ltd 2003. Published by Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ ,
UK and 350 Main Street, Malden, MA 02148, USA.
1394 F. Bowen and K. Blackmon
employees to be a source of change, creativity, and innovation (Frohman, 1997).
Organizational voice describes the voluntary expression of people’s views to influence
organizational actions (Banerjee and Somanathan, 2001; Hirschman, 1970).
Whilst organizational voice can be a powerful source of organizational change,
voice is voluntary and employees are often reluctant to speak out. Organizational
members must therefore choose whether to speak freely and honestly when they
express ideas and opinions on organizational issues. The opposite of voice – orga-
nizational silence – results when people cannot contribute freely to organizational
discourse. As Morrison and Milliken (2000, p. 706) point out, ‘many organizations
are caught in an apparent paradox in which most employees know the truth about
certain issues and problems within the organization yet dare not speak that truth
to their superiors.’
When will people choose organizational voice and when will they choose orga-
nizational silence? As shown in this special issue of the Journal of Management Studies,
interest in voice and silence is on the increase. Further, as an organizational con-
struct, voice is becoming more refined. Morrison and Milliken (2000) have
reviewed the collective dimension of organizational voice and silence at the orga-
nizational level, while other authors, such as Miceli and Near (1992), Ashford et
al. (1998), and Zhou and George (2001), have identified individual attributes asso-
ciated with voice and silence.
In this paper, we use the ‘spiral of silence’ theory developed by Noelle-Neumann
(1974, 1985, 1991) to explain the self-reinforcing collective dynamics of silence
and voice in workgroups. Research on selling issues to top management suggests
that employees will engage in voice when they believe that it will be effective,
relatively low-risk, and where top management is perceived as more supportive
and open (Dutton and Ashford, 1993; Dutton et al., 2001). The spiral of silence
theory, originally developed in the field of public opinion, argues that people scan
their environment to determine the dominant opinion, and express the majority
opinion more readily than a minority one. We apply this theory in the organiza-
tional context, arguing that employees will not use voice unless they are likely to
have support from their co-workers. If they are not confident of support from their
peers, or think that resistance to voice is likely, they will choose silence or dishon-
est responses. Thus, we propose that the choice between voice and silence is largely
determined by the ‘climate of opinion’ in one’s workgroup. This approach to com-
munication within workgroups emphasizes the roles played by the dominant
opinion on an issue, as well as social cohesion, trust and attachment towards the
workgroup.
Although many factors may affect organizational voice and silence, in this paper
we focus on the role of diversity in the dynamics of organizational voice and
silence. We augment Noelle-Neumann’s collective, horizontal, spiral of silence
with an individual, vertical spiral of silence. Managing an increasingly diverse
workplace is a key issue for managers and organizations (e.g., Milliken and Martins,
© Blackwell Publishing Ltd 2003
Spirals of Silence 1395
1996; Tsui et al., 1992). Whilst diversity’s effects on group processes constitute a
vast field of study, within workgroups, we focus on research about the effect of
diversity on workgroup communications and on organizational performance. An
individual whose demographic attributes, social or personal identity differ from
the group norm may not feel able to talk about important aspects of their per-
sonal identity. This may initiate an individual spiral of silence, where they feel
inhibited in making meaningful contributions to the group and in turn the orga-
nization. Small personal silences are thus reinforced and subsequently escalate to
inhibit voice on wider organizational issues. In this paper, we will focus specifically
on how individuals’ choices about disclosing ‘invisible’ personal identities may
spiral outwards to affect organizational voice and silence. People who differ in
‘invisible’ ways such as sexual orientation must choose whether to speak out or
remain silent about their difference. Thus, in this second spiral of silence, employ-
ees who initially remain silent about their personal identities may fail to exercise
organizational voice even if they have valuable contributions that they could
make. We argue that managing ‘invisible diversity’ may be just as crucial as ‘visible
diversity’ such as sex or ethnicity for workgroup communication and hence orga-
nizational performance, but provides additional challenges since individuals may
chose to conceal or evade rather than speak up about their difference (Clair et al.,
2002).
In this paper, we apply the spiral of silence theory together with research on
the effects of diversity on workgroup communications to derive some preliminary
propositions about invisible difference might affect organizational silence and
voice. We focus on sexual orientation, a form of ‘invisible’ diversity, to identify an
agenda for empirical research on silence and diversity. We conclude by describing
an agenda for empirical research and the practical implications of the spiral of
silence framework.
Personal Willingness to
opinion express
opinion
Perception of Assessment
dominant of likely future
opinion opinion
Media Interpersonal
opinions
People’s willingness to express their opinions is influenced not only by their own
personal opinions, but also by their external environment, particularly what they
perceive as the prevailing ‘climate of opinion’. When they are not sure that they
agree with the majority, people are reluctant to express their opinions. They use
their social environments, including public opinion and the judgements of others,
as a frame of reference in interpreting new information and to assess public
opinion. For example, people use opinion polls on probable voter turnout and elec-
tion outcomes in deciding for whom to vote (Scheufele and Moy, 2000). This can
lead to misleading results, for example when in pre-election polls, people report
support for the candidate that they think public opinion supports, rather than their
own choice.
The dominant public opinion exerts control over individuals through the threat
of isolation for deviance. To avoid social isolation, people try to find out both
what the majority opinion is and whether various opinions are increasing or
decreasing in popularity (Noelle-Neumann, 1974). When people perceive that
they share the dominant opinion they will speak out, strengthening this position,
whilst those who perceive that they hold the minority opinion will become more
silent, diminishing their position (Gonzenbach et al., 1999, p. 284). Over time, this
leads to one opinion becoming predominant at the societal level (Schedfele et al.,
2001).
Noelle-Neumann (1991) argued that for spiralling to occur people must perceive
a threat of isolation and they must fear isolation. This is more likely when issues
have a moral or value-laden component, when opinions or attitudes are changing
over time, and when the media take an identifiable position in the conflict,
for instance the ‘don’t ask, don’t tell’ ban on homosexuals in the US military
(Gonzenbach et al., 1999). This leads to a dynamic and self-reinforcing process,
as shown in Figure 2, since:
Time
Individuals tend to publicly express their opinions and attitudes when they per-
ceive their view to be dominant or on the rise. In contrast, when people sense
their view is in the minority or on the decline, they become cautious and silent.
(Scheufele and Moy, 1999)
Even as these other voices point to the omissions and errors and the need for
inclusiveness, the dominant group refuses to hear the message and continues
with the procession. The real issue for the others is getting truly heard, rather
than simply ‘added on’. (Nkomo, 1992, pp. 488–9)
Where individuals who differ from the majority fear this marginalization, they may
avoid speaking up on a wide range of organizational issues. Given the threat and
fear of isolation experienced by some individuals, the conditions for the escalation
of organizational silence in Noelle-Neumann’s spiral may be in place.
We argue that the threat of isolation will be particularly powerful for members of
an ‘invisible’ minority. ‘Invisible’ diversity has been less researched in the man-
agement area than visible forms such as gender or race/ethnicity (Clair et al.,
2002). In this paper, we focus on gay, lesbian, and bisexual employees, whose sexual
orientation we characterize as ‘invisible’ because sexual orientation is not neces-
sarily observable compared with other demographic characteristics such as gender,
age, and race/ethnicity that are usually observable and readily detectable
(Milliken and Martins, 1996; Welch, 1996).
Because sexual orientation is essentially invisible until made visible, gay, lesbian,
and bisexual (GLB) individuals must decide whether to disclose their sexual ori-
entation at work (Day and Schoenrade, 1997), a choice that others may not have
the option to make (Clair et al., 2002). Like members of other non-dominant
groups, they cannot always predict whether they will be accepted by fellow group
members, and may find it difficult to predict the reactions of colleagues, managers,
and other organizational members (Mintz and Rothblum, 1997). Some GLB
employees will always maintain ‘invisibility’ and others will always maintain ‘visi-
bility’; however, most GLB employees will carefully assess the prevailing organi-
zational climate before disclosing this potentially risky or even dangerous aspect
of personal identity. In such conditions, they may choose not to disclose their
sexual orientation until they have gauged whether the local climate of opinion is
likely to be supportive or not.
When GLB members are still invisible, they can sample workgroup opinion as
though they were heterosexual, since group members often behave in the presence
of ‘invisible’ outgroup members in ways that they would not if they were visible.
They can use this information to predict whether co-workers will be supportive,
hostile, or neutral if they ‘come out’. For example, a gay prison service officer
reported that:
In terms of lesbian identity, the university culture was one of silence, and that
was the model I followed. I found it particularly awkward not to do so because
everyone at work knew me as straight and to change this perception would take
what I thought to be a dramatic act. An announcement or something: ‘Excuse
me but . . .’ (Mintz and Rothblum, 1997, p. 7)
Since individuals may feel both the threat and fear of rejection by their workgroup
if they decide to come out, some employees will chose to come out at work, while
others prefer to pass or evade. Thus we argue, based on Noelle-Neumann’s spiral
of silence (see Figure 1), that a key part of this decision for individuals is their per-
ception of the prevailing climate of opinion among their peers. This leads to our
first proposition:
Had I been a better utilitarian, I would have weighed the costs as well as the
benefits of being ‘in the closet’; as it was, I had not even considered the toll
which was exacted by my self-imposed ‘privacy’. How many other gay men, I
wondered, went to work each day representing themselves as something
they were not . . . Were there not likely to be organizational implications for
such an intolerant work environment – if not relating to performance, at least
with respect to commitment, and thereby absenteeism and turnover? (Dunn,
1994)
© Blackwell Publishing Ltd 2003
1404 F. Bowen and K. Blackmon
Nondisclosure was found by Day and Schoenrade (1997, p. 150) to be associated
with lower organizational commitment, because workers do not communicate
their true personal and social identities; lower job satisfaction, because of inhib-
ited natural social interactions; role ambiguity, through less access to information
about the work role through casual conversations; role conflict, through giving up
a part of self-identify; and job stress, through the expenditure of emotional energy
in keeping orientation secret. For example, one respondent quoted in Friskopp and
Silverstein (1995, p. 312) says that:
If you’re closeted at work, people don’t get to know you, so you become alien-
ated from your peer group, and you don’t really like your experience there. In
the closet, you get paranoid. You lose confidence, and it’s much more difficult
to succeed.
‘I cannot speak
credibly about Repression of organizational voice
important issues’
In the station, they like to take care of you. They like to know what you get up
to on your days off; where you go on holiday. At a social function, they like to
see you there with a partner. And if you’re gay, it’s not that easy. You just can’t
bring along a boyfriend, especially if you’re not out. (Cunningham, 2002, p. S3)
When group members are able to express their personal identity freely rather than
trying to manage disclosure, they will be able to engage in a higher level of social
exchange. At the group level, individuals frequently engage in face-to-face inter-
actions (Pelled, 1996). Communications between individuals working together
daily are the small events that make up much of the information sharing and deci-
sion making at work (Larkey, 1996, p. 466). Intra-group communication and inte-
gration is especially important where individuals work together on a problem or
set of problems, and the problem belongs to the whole group (Webster and Hysom,
1998). Further, although people are usually more likely to trust other members of
their own social group (Williams, 2001), familiarity and continued interaction with
someone in their work-related roles can lead to decategorization of people with
stigmatized aspects of personal identity.
However, when sharing information about one’s personal life might lead to
unwanted disclosure of one’s sexual identity, it becomes problematic and individ-
uals will usually try to minimize social exchange in order to maintain their iden-
tity as heterosexual or even asexual.
But at that time, I was working in a paediatrics clinic with very conservative co-
workers. I had heard antigay comments jokes and comments from secretaries,
lab techs, social workers, psychologists, and physicians in this setting over the
© Blackwell Publishing Ltd 2003
1406 F. Bowen and K. Blackmon
years. . . . It did not feel safe to come out at work, so I led a double life for about
three years. I gradually stopped any interaction with co-workers, worked
through lunch breaks, and made excuses for the occasional after work social
gatherings. (Eliason, 1997, p. 43)
On the other hand, when social exchange is low, instrumental exchange may be
low, resulting in a lower level of communication and perceived efficacy. This may
lead to the vicious cycle suggested by Elsass and Graves (1997, p. 951):
Homophily (the preference for similar others) tends to exclude members of out-
groups from opportunities (Pettigrew and Martin, 1987; Thomas and Alderfer,
© Blackwell Publishing Ltd 2003
1408 F. Bowen and K. Blackmon
1989), including the informal networks that provide information and political
contacts (Connelly, 1994; Cox, 1993; Ibarra, 1993; Pettigrew and Martin, 1987).
Therefore, when GLB employees are less integrated with other members, they may
not learn critical information relevant to their work (Day and Schoenrade, 1997),
and miss out on information about opportunities such as promotions and jobs
(Meyerson, 2001; Taylor and Raeburn, 1995). Even when this exclusion is not
intentional, it may result from unconscious avoidance (Gaertner and Dovidio,
1986). Further, even when instrumental exchange is maintained between members
of the ingroup and outgroup, social exchange and support may be limited and
therefore the person is less effective. Finally, if people keep their personal
identity invisible, there are fewer chances to be identified and invited to join
external networks, whilst a visible identity may make an individual distinctive
and therefore more likely to be noticed by members of an external network who
use that as the basis for selection (Mehra et al., 1998). Thus, ‘coming out’ may
improve one’s effectiveness at one’s job, and thus the esteem of fellow workgroup
members.
Contributions
We began this paper by introducing the spiral of silence. We then considered how
individual differences within workgroups might interact with this general spiral of
silence. In particular, where workgroups are heterogeneous, individual sources of
difference can create the potential for valuable contributions but also disrupt work-
group cohesion and communication. We argue in this paper that the effects of
diversity on communications between the focal individual and other individuals in
the workgroup, between the focal individual and external networks, and among
© Blackwell Publishing Ltd 2003
1410 F. Bowen and K. Blackmon
the group as a whole, have the potential to either enable voice or silence of
members of minorities. These individuals’ contributions are potentially most valu-
able, as they are drawn from outside mainstream experience.
We next considered sexual orientation as a source of ‘invisible’ difference, which
workgroup members can choose to reveal and risk social isolation or other nega-
tive effects, or choose to conceal or evade. We then considered the effects of sexual
orientation on the decision to reveal challenging and risky aspects of one’s per-
sonal identity, as the potential seed for a second spiral of silence. Disclosure has
important implications for communications, particularly social exchange, task
communication and instrumental exchange, and ultimately on organizational
voice. Unlike gender and race, which are generally – although not always – observ-
able, sexual orientation is ‘invisible’. The presence of a gay man or lesbian within
a workgroup may be invisible unless that person chooses to disclose his or her
orientation. Thus, sexual orientation offers additional dimensions of diversity to
explore.
Additionally, previous studies have tended to focus on a single aspect of sexual
orientation – concealment, disclosure and discovery of sexual orientation (e.g.,
Clair et al., 2002; Hall, 1989) – rather than their broader implications in the orga-
nizational context that we consider here. Sexual orientation is an issue of growing
importance for organizations, and holds significant moral and status implications.
Besides contributing to the research on sexual orientation within organizations,
sexual orientation has become an important source of employee and customer
diversity, as more people self-identify as gay, lesbian, or bisexual. As a result, orga-
nizations are beginning to address sexual orientation issues. By 2001, 294 Fortune
500 companies included sexual orientation in their equal opportunities policies,
up from 255 in 2000 (Human Rights Campaign, 2001). Along with this, the degree
of legal protection accorded to sexual orientation in the workplace is increasing,
both in scattered locations in the United States (Human Rights Campaign, 2001),
and in Europe under new European Union (EU) law. Further, sexual orientation
may serve as a bellwether for other forms of diversity. For example, a recent study
of high-tech metropolitan areas by Florida and Gates (2001) found that attitudes
towards GLB issues act as an indicator of a progressive and diverse environment
for all workers, including women and people of colour.
In the broader sense, this research may contribute to the literature on organi-
zational sexuality (e.g., Hearn et al., 1989).[1] Humphrey (1999) argues that les-
bians and gay men have become the carriers of organizational sexuality, and thus
their experience can thus be used to reflect the meaning of heterosexuality back
to organizations and their members. Heterosexual identity is omnipresent within
organizations yet little remarked upon. Because gay men and lesbians violate
gender norms, this may also reflect back on the gendering of organizations,
another important concern.
Limitations
Firstly, the extent to which our model based on sexual orientation may be gener-
alized to other types of diversity may be limited. Each demographic variable
should be treated as a distinct theoretical construct rather than assuming all types
of diversity to have similar effects (Pelled, 1996). The effects of sexual orientation
will not necessarily be the same as those of other characteristics such as sex,
race/ethnicity or less controversial invisible diversities. Further, people are not
limited to possessing a single demographic characteristic. The interaction between
sexual orientation and other factors such as age, sex, race/ethnicity, and class may
be complex.
Secondly, we can identify at least two main limitations of the present theoreti-
cal model, including limits to the positive effects of personal identity disclosure,
and the negative effects of involuntary exposure. We argued above that the
voluntary disclosure of sexual orientation enhances organizational voice in a
supportive climate. However, whilst a safe, equitable and welcoming work
environment is essential to realizing the benefits of diversity for decision-making,
I don’t want to be a one issue person. Other staff might believe that I’m only
concerned about gay rights, and I have many other interests beyond homosex-
uality. It might hurt my relationships with other department heads within the
hospital and lessen the degree of respect in which I am held . . . My relation-
ships with conservative or fundamentalist staff members would be damaged.
Practical Implications
A key implication from the spiral of silence for the current area of discussion is
that people’s tendency to ‘jump on the bandwagon’ of what they perceive as the
dominant or strengthening opinion may lead minority group members to perceive
a stronger climate of support for dominant group opinions that actually exists.
Further, it highlights the extent to which processes of voice and silence are self-
reinforcing virtuous or vicious cycles. Finally, understanding the spiral of silence
may be useful if interventions by key individuals such as managers or corporate
leaders, or at key points such as early on in the processes, can be identified and
executed. This may prevent ‘pluralistic ignorance’ from disrupting group and orga-
nizational processes related to diversity.
Fair treatment of all organizational members not only makes sense in terms of
human dignity, but also from organizational self-interest in terms of recruitment,
retention, productivity, and marketing potential. The ability to attract, keep, and
make the best use of qualified individuals – irrelevant of whether their attributes
‘fit’ with the dominant majority – is a hallmark of a leading company. However,
it is not enough to stick diverse individuals together and hope for the best – com-
panies must create the conditions for and actively manage group processes,
especially communications, to ensure that diversity is productive, rather than
destructive. Utilizing our framework of two spirals of silence may help identify
interventions to promote open and honest expression of organizational voice
within organizations.
NOTES
*The authors would like to thank the three anonymous referees and the editors of this special issue
for making insightful and constructive comments that have greatly contributed to this paper. An
earlier version was presented at the 2002 Academy of Management Conference in Denver, CO.
[1] We are grateful to our anonymous referees for pointing out this higher-level linkage.
REFERENCES
Ancona, D. G. and Caldwell, D. F. (1992a). ‘Bridging the boundary: external activity and perfor-
mance in organizational teams’. Administrative Science Quarterly, 37, 4, 634–65.
Ancona, D. G. and Caldwell, D. F. (1992b). ‘Demography and design: predictors of new product
team performance’. Organization Science, 3, 3, 321–42.